Issues about sovereignty can be particularly sensitive and divisive, and are often difficult to prepare for and manage.

Very few countries have ever included in a national constitution provision for a deferred referendum on separation of part of the country required to be held within a specified period, such as the Bougainville case represents.

The only examples I know of are France (in relation to New Caledonia, where a referendum must be held by 2018) and Sudan (in relation to South Sudan, where a referendum was held in 2011, about six years after the Sudan Constitution was amended to provide for it).

So Bougainvilleans are a privileged people to have achieved the opportunity to make a decision about their future in this way.

In all three cases – New Caledonia, South Sudan and Bougainville - such provision was included in the national constitution as part of a broader package intended to find ways of ending bitter and violent conflict.

Autonomy was intended to operate in the period before the deferred referendum, in the hope (for some parties in both New Caledonia and Sudan) that it would help resolve divisions before the referendum was held, perhaps leading to a situation where the referendum might not be necessary, or might be deferred, or perhaps contributing to a referenda outcome in favour of continued unity.

Although referenda can help resolve difficult conflicts, they can also carry risks. They can be complex and expensive to run. They can be divisive, in preparation, in conduct, and in implementation of results.

Problems often arise from misleading and divisive campaigns by political leaders on the question in the referendum. Leaders can be under great pressure to attempt to influence the result through manipulation of the process, and intimidation of voters.

Although usually intended to resolve conflict, holding a referendum can contribute to conflict, especially in a country where there are pre-existing ethnic, religious, or other kinds of divisions.

One particular danger is that the outcome of a referendum on a divisive matter leaves a significant minority feeling strongly that the majority vote causes them serious disadvantage. Violent conflict has occurred in the process of implementation of outcomes of referenda in the past 25 years, including in relation to independence referenda – for example, in East Timor and in South Sudan. 

A difficulty associated with an independence referendum is that it involves a major decision on long-term arrangements being made at a particular time, often without adequate information about future circumstances.

For example, Scotland relies heavily on revenue from petroleum resources, which it would have needed to rely upon if its 2014 referendum had resulted in independence. But little more than a year after the referendum, oil prices were about 25% of what they had been at the time of the vote.

A vote in favour of independence where voters had assumed the prosperity of Scotland was assured could have been ill-founded.

So, in preparing for the Bougainville referendum, it will be important to consider both the advantages and disadvantages that can flow from it, learn from experience of referenda held elsewhere and do everything possible to minimise the chance of serious problems occurring.

A starting point is to develop a clear understanding of the referendum arrangements so that, in planning for and managing it, everything possible is done to ensure arrangements work as intended, potential problems are anticipated and contingencies provided for. 

As yet, however, the referendum arrangements are not widely known and understood. Important aspects are often the subject of confusion, uncertainty and misunderstanding.

The reasons for such confusion include most people involved having no experience of referenda, the complex history of the referendum, almost 15 years having elapsed since the Bougainville Peace Agreement was signed, and few people having a clear memory and understanding of what was agreed.

My paper aims to provide information needed for improved understanding of the arrangements. The matters discussed in the paper highlight both the complex origins of the referendum arrangements and the complexity inherent in the arrangements.

These factors contribute significantly to the confusion and misunderstandings about fundamentally important aspects of the arrangements, not only amongst Bougainvilleans but also at the national government level. The extent of confusion and misunderstanding is likely to contribute to tensions and even violence if not addressed through extensive awareness and consultation programs.

The tasks involved in preparation for the referendum will be numerous and in many respects complex. Many will involve consultation and agreement between the two governments.

There will also be a necessity for prior awareness and consultation involving Bougainvillean communities before the Autonomous Bougainville Government engages with the national government in relation to such issues.

Despite the extent of the preparatory efforts required, very little has been done to initiate most of the necessary steps arising from the legal arrangements.

If the referendum is to be held before the end of the five year window (mid-2020), it is becoming increasingly urgent that preparatory work begin soon.

The most obvious practical steps would involve establishing both a joint secretariat and the agency required to conduct the referendum.

Establishing the agency is of particular importance because, as an independent body, there will be much that it can do without the need for approval from either government, subject mainly to the availability of resources necessary to begin its work.